RNA sequencing reveals resistance of TLR4 ligand-activated microglial cells to inflammation mediated by the selective jumonji H3K27 demethylase inhibitor

Persistent microglial activation is associated with the production and secretion of various pro-inflammatory genes, cytokines and chemokines, which may initiate or amplify neurodegenerative diseases. A novel synthetic histone 3 lysine 27 (H3K27) demethylase JMJD3 inhibitor, GSK-J4, was proven to exert immunosuppressive activities in macrophages. However, a genome-wide search for GSK-J4 molecular targets has not been undertaken in microglia. To study the immuno-modulatory effects of GSK-J4 at the transcriptomic level, triplicate RNA sequencing and quantitative real-time PCR analyses were performed with resting, GSK-J4-, LPS- and LPS + GSK-J4-challenged primary microglial (PM) and BV-2 microglial cells. Among the annotated genes, the transcriptional sequencing of microglia that were treated with GSK-J4 revealed a selective effect on LPS-induced gene expression, in which the induction of cytokines/chemokines, interferon-stimulated genes, and prominent transcription factors TFs, as well as previously unidentified genes that are important in inflammation was suppressed. Furthermore, we showed that GSK-J4 controls are important inflammatory gene targets by modulating STAT1, IRF7, and H3K27me3 levels at their promoter sites. These unprecedented results demonstrate that the histone demethylase inhibitor GSK-J4 could have therapeutic applications for neuroinflammatory diseases.

such as NO, prostaglandin E2 (PGE2), cyclooxygenase-2 (COX-2), cytokines (e.g., IL1B, IL6 and TNF-α) and TFs (e.g., NF-κB, IRF, STAT) in microglial cells 6 . From previous studies, the response of macrophages to LPS is believed to proceed through histone modification at specific inflammatory genes, prompting further exploration to address the temporal cascade of epigenetic events and the effects of specific epigenetic inhibitors. A pivotal study showed that upon LPS stimulation, the H3K27 demethylase JMJD3 was rapidly induced and that more than 70% of LPS-induced genes recruited JMJD3 to their promoter region, which is a general hallmark of gene activation 7 . Indeed, JMJD3 is a bona fide mediator of H3K27me1/me2/me3 demethylation, by reprogramming the transcription of genes by recruiting distinct TFs to gene promoters via epigenetic regulation that is involved in pro-inflammatory gene transcription [7][8][9] . These observations raised the possibility that JMJD3 may contribute to demethylation-dependent histone-packaged inflammatory gene expression programs associated with various human diseases.
Recently, a potent and highly specific JMJD3 inhibitor, GSK-J4, was discovered by Kruidenier and colleagues 10 . GSK-J4 reduces pro-inflammatory cytokine production by modulating JMJD3, leading to a reduction in H3K27me3 in LPS-induced macrophage cells 11,12 . Recently, GSK-J4 has been demonstrated to exhibit a potent inhibitory activity against a range of cell lines derived from certain cancers, including T-ALL, B-ALL and glioma 11,12 . These studies empower the mechanistic investigation of how this inhibitor can be used to effectively modulate JMJD3 in microglial cells.
Although GSK-J4 decreases the production of inflammatory cytokines in LPS-induced macrophages 10 , a genome-wide search for GSK-J4 molecular targets in LPS-induced microglial cells has not yet been performed. Therefore, we studied gene array and comparative gene expression analyses from PM and BV-2 microglial cells upon stimulation with LPS, GSK-J4, and LPS + GSK-J4 using massively parallel cDNA sequencing (RNA-seq), which opened the way to unbiased and efficient assays on the transcriptome of any mammalian cell 13,14 . In principle, RNA-seq allows the identification of all expressed transcripts, both protein-coding and non-coding. Furthermore, several studies indicate that next-generation sequencing is more valuable and particularly appropriate to examine the pathogenesis of complex neurodegenerative diseases such as AD 15 . To the best of our knowledge, this is the first genome-wide study of GSK-J4-mediated global gene expression changes in PM and BV-2 microglial cells using RNA-seq analysis.
Here, for the first time, we report that GSK-J4 is a potent modulator of microglial activation. In particular, GSK-J4 treatment resulted in the significant down-regulation of key inflammatory genes in LPS-induced PM and BV-2 microglial cells. Importantly, these inflammatory genes were not affected by GSK-J4 treatment alone. Overall, the results suggested that the synthetic compound GSK-J4 might be an effective therapeutic target with possible research and clinical value. Taken together, these findings establish a role for GSK-J4 in mouse microglia stimulation and justify the further testing of GSK-J4-targeting genes in the treatment of inflammatory brain damage.

Results
Distinct gene signatures were identified during the inflammatory response using RNA-seq analysis in PM and BV2 microglial cells. To determine the high-resolution transcriptome in response to LPS stimulation, we treated PM and BV-2 microglial cells with LPS for 4-h before cDNA library preparation for RNA-seq experiments. The RNA-seq transcriptional analysis was performed using three independent samples (biological replicates) of each treatment in PM. The data from all the experiments (each group) were combined, and the genes whose expression levels significantly differed were identified. We used a 1% false discovery rate (FDR), P ≤ 0.01, and log 2 -fold change ≥1.5 for up-or down-regulation as the criteria for defining the differentially expressed genes. Of note, we previously found that most of the inflammatory response-related genes were up-regulated at a 4-h time point 16,17 . We chose this time point for transcriptional profiling; this time point was also used in other studies 18,19 that investigated the general induction pattern of microglial activation by LPS.

GSK-J4 regulates common and unique transcriptional programs between PM and BV-2 microglial cells.
To further investigate the common and unique GSK-J4 down-regulated genes between the LPS-treated PM and BV-2 microglial cells, we again used RNA-seq data to compare the GSK-J4 down-regulated transcriptome of BV-2 microglial cells with PM. Using a similar approach (see above), we compared the GSK-J4 down-regulated transcripts in LPS-treated BV-2 microglial cells with that of PM. Differential expression analysis clearly revealed that GSK-J4 down-regulated a unique gene set in response to LPS stimulation in PM and BV-2 microglial cells (Fig. 3I) suggesting a substantial number of dissimilarities between the two cell types. GSK-J4 down-regulated 430 genes in PM cells that are not common to BV-2 microglial cells. In contrast,  GSK-J4 down-regulated 32 genes in BV-2 microglial cells that are not common to PM cells (Fig. 3I). The GSK-J4 down-regulated unique and common gene sets for PM and BV-2 microglial cells are presented in Tables 1, 2 and 3. Of the GSK-J4 down-regulated genes, BV-2 microglial cells and PM shared 107 genes following LPS treatment (Fig. 3I). Importantly, this technology allowed us to identify several specific GSK-J4 down-regulated gene families involved in immune responses that were uniquely altered in LPS-treated PM cells. The major GSK-J4 down-regulated genes in PM only included cytokines/chemokines (CCL3, CCL4, CCL5, CCL8, CCL22, CXCL1,  CXCL3, CXCL9, CXCL11, CXCL16, IL1BOS, IL1RN, IL12B, IL13RA, IL15RA, IL15, IL18, IL18BP, IL23A, and  IL27) and IRGs (GBP2B, GBP4, GBP6, GBP9, GBP10, GBP11, IFI35, IFI44I, IFI202B, and IFITM3) ( Table 1). These data suggest that GSK-J4 down-regulated a unique set of genes that is distinct from that of BV-2 microglial cells and may offer new potential GSK-J4 molecular targets in LPS-induced microglial activation.

Differential expression of TFs and signaling pathways modulated through GSK-J4 in LPS-induced PM and BV-2 microglial cells. To further investigate whether the effect of GSK-J4 altered
the key TFs associated with inflammation, we examined multiple families of TFs that were identified after 4-h of LPS stimulation in PM and BV-2 microglial cells. The annotation of the RNA-seq data also revealed that in the presence of LPS stimulation, GSK-J4 also altered the expression of some key TFs, with a log 2 -fold change ≥1.5 and P ≤ 0.01 cut-off values ( Fig. 4A and Supplementary Figure S4A). These TFs, including IRF and STAT, play important roles in neuroinflammatory diseases 23,24 . The following TF families exhibited the most dramatic suppression levels following the GSK-J4 challenge: (1) the IRF group of TFs (IRF7 and IRF9);  Figure S4A), suggesting that GSK-J4 suppression of TF expression is highly selective in PM and BV-2 microglial cells. Next, we conducted a TF motif analysis to assess the GSK-J4-suppressed gene expression in PM and BV-2 microglial cells. We used the Pscan software tool 25 to perform an in silico computational analysis of over-represented cis-regulatory elements within the 5′-promoter regions of coordinately regulated genes. Applying this score to the gene promoters suppressed by GSK-J4 treatment at 4-h in response to LPS stimulation revealed that the putative binding sites for STAT1, STAT1-STAT2, IRF7, and IRF9 were significantly enriched (Fig. 4D,E and Supplementary Figures S4D,E). Next, we analyzed how many down-regulated genes contain a STAT1 and IRF7 binding motif in the promoter sequence. Interestingly, among down-regulated genes we found a significant percentage of the cytokines and chemokines as well as IRGs had a STAT1 (328/537; 61%) and IRF7 (318/537; 59%) binding motif in the promoter  region (from −950 bp to +50 bp), these data are summarized in Fig. 4F and Supplementary Table S1, for PM cells.
We also found a significant percentage of STAT1 and IRF7 binding motifs in the promoter region of the cytokines and chemokines as well as IRGs in BV-2 microglial cells (Supplementary Figure S4F and Supplementary Table S1). In addition to TF motif analysis, we also applied IPA software 26 to identify the target genes that were directly or indirectly activated by the identified TFs in response to GSK-J4 treatment. Importantly, the assessment of  Table 3. Top 50 GSK-J4 down-regulated common genes in LPS treated PM and BV-2 microglial cells. (H) ChIP assay to determine the presence of STAT1 and IRF7 at selected genes. The ChIP-enriched samples were analyzed using quantitative PCR with selected genes primers. STAT1 and IRF7 binding was increased following LPS exposure, though a reduced presence of STAT1 and IRF7 binding was shown at the promoters of the CCL2, CCL7, and CXCL10 genes in GSK-J4-treated BV-2 microglial cells. The graphs represent the mean fold values of enrichment relative to the IgG control from three independent experiments. *P < 0.01 and **P < 0.001 compared with the control.
upstream regulators by IPA similarly revealed that down-regulation of most of the cytokines and chemokines was also directly regulated by the identified TFs, including STAT1 and IRF7 (Fig. 4G, Supplementary Figure S4G and Table 4).
We also hypothesized that STAT1 and IRF7 would also associate with the target gene promoters. Accordingly, we performed chromatin immunoprecipitation (ChIP) studies to map STAT1 and IRF7 occupancy along the target key pro-inflammatory gene promoter regions in LPS-induced BV-2 microglial cells. As shown in Fig. 4H, LPS-induced STAT1 and IRF7 binding and co-treatment with GSK-J4 led to STAT1 and IRF7 binding interference at the promoter site of key pro-inflammatory genes in BV-2 microglial cells. Our results are consistent with a previously published report that showed that STAT1 binds to the promoter region of several inflammatory genes, and that silencing of the JMJD3 gene significantly repressed the expression of STAT-dependent genes 27 . However, we could not identify significant binding sites for STAT3 at the promoter site of GSK-J4-suppressed genes in BV-2 microglial cells and PM. Furthermore, we again performed ChIP studies to map STAT1 and IRF7 occupancy along the GSK-J4 non-target gene promoters. We found that LPS-induced STAT1 and IRF7 binding and co-treatment with GSK-J4 did not lead to a decrease in STAT1 and IRF7 binding at the promoter site of GSK-J4 non-target genes in BV-2 microglial cells (Supplementary Figure S5). Taken together; these findings suggest that STAT1 and IRF7 TFs might be involved in the regulation of GSK-J4-suppressed microglial cell activation.
Effect of GSK-J4 alone on resting PM cells. We wondered whether the effect of GSK-J4 alone altered the genes associated with inflammation in resting PM and BV-2 microglia cells. To address this point, we examined the genome-wide effect of GSK-J4 alone in resting PM and BV-2 microglia cells. Our results revealed that GSK-J4 alone, in the absence of LPS stimulation, also altered the expression of genes, with a 1.5 log 2 -fold and P ≤ 0.01 cut-off value. Surprisingly, genes associated with the LPS response and inflammation (CCL6, CCL8,  CX3CL1, CXCL1, CXCL3, CXCL9, CXCL11, CXCL16, IL12B, IL18BOS, IL18BP, IL19, IL23A, IL27, IRAK1BP1, SOCS1, TNFSF11A, and TNFSF15) were unaffected or were expressed insignificantly. A total of 278 and 3 genes (log 2 -fold ≥ 1.5 and P ≤ 0.01) were up-regulated in the PM and BV-2 microglia cells, respectively, that were treated with GSK-J4 alone at 4-h (Fig. 3A,B). Notably, we observed that heat shock protein (HSP) 1 A and 1B, metallothionein (MT) 1 and 2, RASD family member 2 (RASD2), etc. genes were up-regulated in GSK-J4 stimulated PM at 4-h ( Supplementary Figures S6A-C). HSPs are class of molecular chaperones that function in the regulation of both necrotic cell death and cell survival. Dysregulation of HSP expression has been demonstrated to play critical roles in the pathogenesis of CNS diseases. For example, HSPA1B was induced in the CNS of MS patients and experimental autoimmune encephalomyelitis (EAE) animals, and it is believed that the inhibition of HSPA1B might be a therapeutic target for EAE in MS patients 28,29 . In contrast, other studies showed that overexpression of HSPA1B may have a potential role against brain ischemia via an anti-inflammatory mechanism 30,31 . In addition to HSPA1B, HSPA1A overexpression is likely to be responsible for the protection seen in ischemia 32 . MT-1/MT-2 has been implicated in a wide array of pathological conditions in the brain and in neurodegenerative diseases. In particular, MT-1/MT-2 isoform expression was up-regulated in animal models of MS 33 . RASD2 plays a critical role for the selective toxicity in HD and is suggested to be a potential new target for HD 34,35 . We confirmed with DAVID Bioinformatics Resources GO analysis (FDR 0.05) that GSK-J4-up-regulated transcripts were associated with rhythmic processes, growth, as well as immune system process (Supplementary Figure S6D). A better understanding of the consequences of only GSK-J4-mediated modulation of microglia activation warrants a comprehensive investigation.

Functional and pathway analyses following GSKJ4 treatment in LPS-induced PM and BV-2 microglia cells.
To further explore and functionally classify the GSK-J4 down-regulated genes with LPS stimulation, we again used the DAVID Bioinformatics Resources. Interestingly, we observed that the largest gene groups were involved in the same biological processes, e.g., immune system processes, cell killing, and multi-organism processes (Fig. 5A,B). To determine the potential biological pathways of the GSK-J4-down-regulated genes in the LPS-induced PM and BV-2 microglia cells, we utilized the PANTHER  classification system, version 9.0 36 . The major categories of the biological pathways were inflammation mediated by chemokines and cytokines, interleukin and integrin signaling pathways (Fig. 5C,D). To corroborate these functional findings, we analyzed the influence of GSK-J4 on molecular signaling networks in LPS-induced PM and BV-2 microglia cells using IPA 26 . Additionally, GSK-J4 down-regulated gene sets revealed signaling networks related to antimicrobial activity, inflammatory response, and infectious diseases. Particularly, three TFs, including IRF7, STAT1 and STAT2, were identified as the main modulator genes (Fig. 5E,F). Together these data clearly implicate that GSK-J4 may limit the inflammatory response in microglial cells.

Confirmation of GSK-J4 down-regulated genes by qRT-PCR in PM and BV-2 microglia cells.
Several genes that were identified by RNA-seq analysis as differentially regulated were subjected to validation through real-time qRT-PCR using GAPDH as a reference gene. Most were selected for validation according to the distinct effects of GSK-J4 on the LPS-affected genes. To measure gene expression, mRNA was reverse transcribed into cDNA using the Prime Script TM Reverse Transcriptase (Takara Bio Inc., Shiga, Japan) and the qRT-PCR assays were repeated several times using at least 3 mRNA preparations from independent biological experiments. The results are expressed as the fold change relative to the control levels. We found that in almost all cases, there was very good agreement between the RNA-seq and the qRT-PCR results in terms of the direction of change as well as its magnitude. The expression levels of the mRNA of nine genes selected for verification, and the RNA-seq expression pattern was confirmed for eight genes (CCL2, CCL7, CCL9, CXCL10, IL1RN, IRG1, IRF7 and IL6; Figs 2 and 6A,B) in PM and BV-2 microglia cells however, one was non-significant (data not shown) in the qRT-PCR analysis compared with the RNA-seq experiments.

GSK-J4 inhibits inflammatory genes through H3K27 demethylation in BV-2 microglia cells.
To further elucidate the biological relevance of H3K27 demethylation, we performed ChIP studies on GSK-J4 treated BV-2 microglia cells in the presence or absence of LPS. Initially, we targeted the genes from RNA-seq datasets that were significantly down-regulated in the GSK-J4 treated samples but normally up-regulated in the LPS-induced samples (CCL2, CCL7, and CXCL10). We selected the important key pro-inflammatory mediators. Using ChIP assays, we observed that that GSK-J4 prevented the LPS-induced loss of H3K27me3 levels at the promoter site of key pro-inflammatory genes (Fig. 6C), suggesting that global gene expression changes in microglia cells after GSK-J4 treatment by inhibiting H3K27me3 demethylation. We also confirmed the reduced expression of these genes and found that they were down-regulated in the GSK-J4-treated samples (Fig. 2).

GSK-J4 inhibits pro-inflammatory cytokine expression in mouse brain microglia. The in vivo
effect of GSK-J4 on neuroinflammation was examined in an established model of neuroinflammation 37 . After LPS challenge (1 mg/kg), the mice received an intraperitoneal injection of GSK-J4 (1 mg/kg). LPS administration (1 mg/kg) significantly elevated the expression of pro-inflammatory genes in adult microglia (Fig. 7). More importantly, subsequent injection of GSK-J4 significantly suppressed the expression of key LPS-inducible inflammation and immunity-related genes, including CCL2, CCL3, CCL4, CCL12, IRF1, IL1A, IL1B, and IRG1 in the adult microglial cells (Fig. 7). However, it should be noted that the CXCL10 gene was not suppressed by GSK-J4 treatment. Additionally, these pro-inflammatory mediators were not affected by only GSK-J4 injection in adult microglia.

Discussion
JMJD3 has emerged as a crucial epigenetic regulator governing the assembly of histone demethylation-dependent chromatin complexes that regulate inflammatory gene expression in macrophages 7,38 . A potent and highly specific JMJD3 inhibitor, GSK-J4 has been shown to disrupt inflammation and cancer 11,12 . For example, GSK-J4 can inhibit the transcription of a plethora of pro-inflammatory genes in LPS-induced macrophages, including crucial inflammatory genes, such as TNF-α 10 . Additionally, other studies reported even wider prospective applications for GSK-J4, such as in attenuating brainstem glioma or T-cell acute lymphoblastic leukemia (T-ALL), suggesting that GSK-J4 may have anti-inflammatory as well as anti-cancer activities 12 . However, none of these studies addressed the effects of GSK-J4 at the genome-wide expression level in microglial cells. Because previous studies have demonstrated that PM showed a unique molecular expression profile that was different from the profile in BV-2 microglia cells 17, 39 , we examined both PM and BV-2 microglial cells as a model of inflammation, which is one of the major experimental uses of microglia. For the first time, in the present study, we identified a prominent transcriptional response in resting as well as LPS-induced microglial cells after GSK-J4 treatment using RNA-seq analysis. This unbiased profiling approach revealed the importance of GSK-J4 in the regulation of key inflammatory genes involved in the establishment of innate immunity in microglia. Our RNA-seq data revealed that GSK-J4 repressed the expression of an important subset of pro-inflammatory genes, cytokines/chemokines as well as IRGs, including CCL2, CCL7, CCL9, CCL12, CXCL10, IL1A, IL1B,  IL1RN, IL6, IL15, GBP2, GBP3, GBP5, GBP7, GBP9, GBP10, GBP11, IFIT1, IFIT2, IFIT3, IFIT3B, IFIH1, IFI35,  IFI44, IFI47, IFI203, IFI204, IFI205, IRGM1, IRGM2, ISG15, ISG20, MX1, MX2, OASL1, OASL1B, OASL1G,  USP12, USP18, USP21, USP25 and ZBP1 (Fig. 3C-H). These cytokines/chemokines are also referred to as inflammatory cytokines, and their excessive production has been associated with disease progression and severe inflammation pathologies, including trauma, ischemic injury and MS 40 . CCL2 and CCL7 are potent chemoattractants for monocytes/macrophages and are highly expressed in microglia, astrocytes and other inflammatory cells during MS 41,42 . Additionally, Conductier et al. reported that CCL2 plays a crucial role in neuroinflammatory diseases, and considered it as a target in the treatment of neuroinflammatory disorders 43 . One of the mechanisms to activate microglia is the expression of CD40; it is believed that the induction of CD40 is critical for a productive immune response. Ultimately, increased CD40 expression coupled with the secretion of CCL2 and CCL7 results in exacerbation of neuroinflammation 44,45 . The expression of CXCL10 or interferon gamma-induced protein 10 (IP-10) has been observed during several neurodegenerative diseases and plays a crucial role in T-cell-mediated inflammation in the CNS 46 . CXCL10 also has a well-established role in inflammatory demyelinating diseases, such as MS, through the destruction of the myelin sheath or neurons by facilitating leukocyte trafficking in the brain 47 . Thus, the down-regulation of numerous inflammatory genes (CCL2, CCL7, CCL9, CCL12, CXCL10,  IL1A, IL1B, IL1RN, IL6, IL15, GBP2, GBP3, GBP5, GBP7, GBP9, GBP10, GBP11, IFIT1, IFIT2, IFIT3, IFIT3B,  IFIH1, IFI35, IFI44, IFI47, IFI203, IFI204, IFI205, IRGM1, IRGM2, ISG15, ISG20, MX1, MX2, OASL1, OASL1B, OASL1G, USP12, USP18, USP21, USP25 and ZBP1) through GSK-J4 would have great utility in reducing neuroinflammation in the CNS.
It is well-known that IRFs and STATs are important TFs involved in the regulation of inflammation disorders, including neurodegenerative diseases. We showed here that GSK-J4 repressed the expression of an important subset of LPS-inducible TFs, especially IRF7, IRF9, STAT1, and STAT2 (Fig. 4A). STATs have been implicated in several CNS pathologies. Due to its critical role, the STAT family is suggested to be one of the most extensively studied targets in inflammation 34 , and inhibition of STAT activity is likely to function as a potent therapeutic strategy 48 . IRFs are a family of TFs that are involved in inflammatory diseases 49 . Type 1 IRFs have well-characterized pro-inflammatory and anti-viral roles in neuroinflammation. Indeed, IRF7 is a significant regulatory factor in the development of demyelination diseases in the CNS, such as MS and EAE 33 , whereas IRF9 is important in injury-induced type 1 IRF signaling, which regulates inflammatory responses in the CNS 50 . Furthermore, GSK-J4 also inhibited the expression of a wide group of other IRGs in LPS-induced PM and BV-2 microglial cells. Thus, The most highly represented biological pathways of the GSK-J4 down-regulated genes in PM (P ≤ 0.01, and log 2fold change ≥1.5) and BV-2 microglial cells (P ≤ 0.01, and fold change ≥1.5). (E,F) Ingenuity ® Bioinformatics pathway analysis of gene networks displaying interactions between infectious diseases, antimicrobial and inflammatory response-related genes that were down-regulated by GSK-J4 at 4-h after LPS stimulation. Genes in white circles were not in our DEG dataset but were inserted by IPA because these genes are connected to this network. The activity of molecules highly connected to this network, namely IRF7, STAT1 and STAT2 (hubs), was assessed using the IPA molecule activity predictor.
the down-regulation of STAT1, STAT2, IRF7 and IRF9 through GSK-J4 could inhibit neurodegenerative diseases as well as brain inflammation. Finally, the results achieved by the real-time RT-PCR analysis of CCL2, CCL7, CCL9, CXCL10, IL1RN, IRG1, IRF7 and IL6 (Fig. 6A) illustrate the essential down-regulated expression of the abovementioned mRNAs in GSK-J4-treated PM and BV-2 microglial cells when compared to the control. Moreover, ChIP studies confirmed that GSK-J4-down-regulated genes prevented the LPS-induced loss of H3K27me3 as well as interfered with the STAT1 and IRF7 binding at the promoter site of key pro-inflammatory genes in BV-2 microglial cells (Fig. 6D).
One of the most striking features is that GSK-J4 significantly suppressed the expression of previously unidentified inflammatory genes that are induced by LPS in PM and BV-2 microglial cells (Fig. 3C,D). MMPs play a pivotal role in neuroinflammation and neurodegenerative disorders, and MMP13 has been found to be activated in brain tissue 51,52 . Orphan G protein-coupled receptor 84 (GPR84) expression is restricted to microglia and observed in different pathological conditions, including EAE and endotoxemia 53 . PELI1 plays a critical role in microglia activation during EAE pathogenesis and is suggested to be a potential new target for MS therapy 54 . Accordingly, our current results showed down-regulation of MMP13, GPR84, and PELI1 expression (Fig. 3C,D); the inhibition of these genes by GSK-J4 could underlie the potential benefits in treatment of neuroinflammatory diseases. However, the mechanism by which GSK-J4 inhibits key inflammatory genes requires further study.  IL1RN, IRG1, IRF7, IL6, CCL2, CCL7, CCL9, and CXCL10 AND IL1RN,  IRG1, IRF7, and IL6 genes were significantly down-regulated in the GSK-J4-treated PM and BV-2 microglial cells, respectively. Gene expression was normalized to the GAPDH transcript levels. *P < 0.01 and **P < 0.001 compared with the control. The data represent three independent biological experiments. (C) ChIP assay to determine the presence of H3K27me3 at selected genes. The ChIP-enriched samples were analyzed using quantitative PCR with selected gene primers. The H3K27me3 levels were decreased following LPS exposure while an induced presence of H3K27me3 was shown at the promoters of the CCL2, CCL7, and CXCL10 genes in GSK-J4-treated BV-2 microglial cells. The graphs represent the mean fold values of enrichment relative to IgG control from three independent experiments. *P < 0.01 and **P < 0.001 compared with the control.
In the absence of LPS stimulation, the treatment of PM with GSK-J4 did not have an impact on LPS-induced inflammatory gene expression. It is interesting to note that crucial inflammatory genes, including CXCL2 and TNF-α as well as other inflammatory and immunity-related genes, such as CCRL2, CSF1, DUSP16, IGSF6, IGSF8, IL17RA, IFNB1, JUNB, REL, RELB, TNFAIP3, TNFSF9, and TNIP3, were unaffected by GSK-J4 in PM and BV-2 microglial cells (Supplementary Figure S3). This specificity and anti-inflammatory potential of GSK-J4 was validated using qRT-PCR analysis (data not shown). In contrast, Kruidenier et al. demonstrated that GSK-J4 is a potent inhibitor of TNF-α production in macrophages 10 . Importantly, we observed that the prominent TFs STAT1, STAT2, IRF7, and IRF9 were suppressed by GSK-J4, although surprisingly, GSK-J4 had no effect on the master TFs NF-kB (REL and RELB) or AP-1 (ATF-3 and JUNB) in PM. It has been demonstrated that the NF-kB pathway is a critical player in the regulation of TNF-α and CXCL2 expression in macrophages 55,56 . Furthermore, other reports have shown that both the AP-1 and NF-kB pathways prominently regulate TNF-α gene expression 57 . Therefore, it seems likely that the LPS-induced induction of TNF-α and CXCL2 transcription depends on NF-kB or AP-1, rather than the STAT1, STAT2, IRF7, and IRF9 transcriptional pathways in microglia. Furthermore, we confirmed that GSK-J4 did not lead to a decrease in STAT1 and IRF7 binding at the TNF-α and CXCL2 gene promoters in BV-2 microglial cells (Supplementary Figure S5). A better understanding of the Figure 7. In vivo effect of GSK-J4 on pro-inflammatory responses in LPS-challenged mice. ICR mice (n = 5 for each group) were treated with LPS (1 mg/kg) following GSK-J4 injection (1 mg/kg) and brain microglia were collected at 4-h to determine LPS-induced gene expression by qRT-PCR. The CCL2, CCL3, CCL4, CCL12, IRF1, IL1A, IL1B, and IRG1 genes were significantly down-regulated in the GSK-J4-injected adult microglial cells. Gene expression was normalized to GAPDH transcript levels. Each point represents data from an individual mouse, all values shown as mean ± S.E.M. *P < 0.01, **P < 0.001 and ns is non-significant versus all other groups; calculated by two-way ANOVA Tukey's HSD post-hoc test.
consequences of LPS-induced cytokines/chemokines production modulated by GSK-J4 in microglial cells warrants a comprehensive investigation.
Interestingly, a recent report demonstrated that systemic GSK-J4 treatment decreased the development of EAE in mice 58 . However, the identification of the GSK-J4 molecular target in LPS-induced microglia has not yet been examined using RNA-seq analysis. Collectively, our RNA-seq data raised the possibility that GSK-J4 may play a potential role in the treatment of neuroinflammatory diseases, possibly through its inhibition of microglia and the following inflammatory responses in the CNS. We also showed that GSK-J4 controls important inflammatory gene targets by modulating STAT1, IRF7, and H3K27me3 levels at their promoter sites. Nevertheless, further extensive in vivo research is needed to investigate the anti-inflammatory effect of GSK-J4 and its underlying mechanism, which may finally result in the development of effective and safe anti-inflammatory drugs.

Conclusion
On the basis of RNA-seq data we speculated that GSK-J4 triggered global changes in resting and LPS-induced microglial cells transcriptomes, targeting inflammatory diseases of the CNS. The findings suggest for the first time that GSK-J4 selectively inhibits the expression of several immune and inflammation-related genes, including cytokines, chemokines, interleukins, interferons and TFs, as well as novel inflammatory targets, to exert its anti-inflammatory functions. Moreover, GSK-J4 could be targeted by epigenetic-focused drug discovery, as it may have therapeutic applications in inflammation-mediated neurodegenerative diseases. Adult mouse microglia were prepared from 5-6-weeks-old ICR mice as previously described 60 with minor modifications. Briefly, mouse brains (excluding the brain stem but including the cerebellum) were dissected out of the skull, and blood vessels and meninges were carefully removed. Then, the mouse brains were cut into small pieces (1-2 mm), vigorously triturated to obtain single cell suspensions, and digested using a Neural Tissue Dissociation Kit-Postnatal Neurons (Miltenyi Biotec, Germany, 130-094-802). Myelin debris was removed using modified protocols from Miltenyi Biotec's, Germany (Myelin Removal Beads II, 130-096-731). After myelin removal, cells were stained with PE-conjugated anti-CD11b antibodies (Miltenyi Biotec, Germany, 130-093-634) followed by incubation for 15 minutes with anti-PE magnetic beads. Magnetically tagged CD11b + cells were then isolated using MS columns according to the Miltenyi Biotec MACS protocol. For immunocytochemistry, purified microglial cells were seeded in 4-well plates. The purity of the primary microglia was 90-91% as determined by immunocytochemistry with antibodies against CD11b 61 (Supplementary Figure S1B Immunocytochemistry. Microglia cells were seeded onto coverslips in 4-well plates. The cells were washed with PBS, fixed with 4% paraformaldehyde for 15 min and then permeabilized with cold methanol for 5 min. After blocking with 5% BSA in PBS for 1 h, the cells were incubated with primary antibody overnight at 4 °C with CD11b monoclonal antibody (1:200, Abcam, Cambridge, U.K) followed by incubation with an appropriate secondary donkey anti-rabbit IgG antibody (Thec Jackson Laboratory, West Grove, PA) for 1 h at room temperature. After incubation, the cells were washed with PBS three times and then mounted with 4′,6-diamidino-2-phenylindole (DAPI)-containing mounting solutions (Vectashield, Vector Laboratories, Burlingame, CA) and imaged with an immunofluorescence microscope (Nikon, Tokyo, Japan).

Total RNA isolation and cDNA library preparation for transcriptome sequencing (RNA-seq).
Total RNA was extracted using RNAiso Plus (Takara Bio Inc., Shiga, Japan) and a QIAGEN RNeasy ® Mini kit (QIAGEN, Hilden, Germany). PM or BV-2 microglial cells were completely lysed using RNAiso Plus, and then 200 μl of chloroform was added. The tubes were then inverted for 5 min. The mixture was centrifuged at 12,000 × g for 15 min at 4 °C, and the upper phase was placed into a new tube. A 600 μl volume of 70% ethanol was added, and the mixture was applied to an RNeasy mini column. The column was washed with wash buffer. To elute the RNA, RNase-free water (30 μl) was added directly onto the RNase mini column, which was then centrifuged at 12,000 × g for 3 min at 4 °C. To deplete ribosomal RNA (rRNA) from the total RNA preparations, a RiboMinus Eukaryote kit (Life Technologies, Carlsbad, CA) was used according to the manufacturer's instructions. RNA libraries were created using a NEBNext ® Ultra ™ directional RNA library preparation kit for Illumina ® (New England Biolabs, Ipswich, MA). The obtained rRNA-depleted total RNA was fragmented into small pieces using divalent cations at elevated temperatures. First-strand cDNA was synthesized using reverse transcriptase and random primers, and second-strand cDNA synthesis was then performed using DNA polymerase I and RNase H. The cDNA fragments were processed using an end-repair reaction after the addition of a single ' A' base, followed by adapter ligation. These products were purified and amplified using PCR to generate the final cDNA library. Differentially expressed gene analysis using RNA-seq data. FASTQ files from RNA-seq experiments were clipped, trimmed of adapters, and the low-quality reads were removed by the Trimmomatic 62 Quality controlled FASTQ files were alignment to Mus musculus UCSC mm10 reference genome sequence using the STAR (version 2.5.1) aligner software 63 . To measure differential gene expression, DESeq. 2 64 with the default parameters was used. A subset of condition-specific expression was defined as showing a log 2 fold change ≥1.5 and P ≤ 0.01 in expression between controls, GSK-J4, LPS, and LPS + GSK-J4 treated samples. The RNA-seq experiments were visualized using HOMER 65 after custom tracks were prepared for the UCSC Genome Browser (http://genome. ucsc.edu/). The acquired data were deposited in the Gene Expression Omnibus database under dataset accession no. GSE79898, GSE80304 and GSE89817.
Quantitative real-time RT-PCR (qRT-PCR). The reverse transcription of the RNA samples was performed as previously described 17 using 2 µg of total RNA, 1 µl of oligo (dT) primer (per reaction) and a Prime Script 1st strand cDNA Synthesis Kit (Takara Bio Inc., Shiga, Japan). The oligo (dT) primer and RNA templates were mixed and denatured at 65 °C for 5 min and then cooled for 2 min on ice. Prime Script buffer (5x), RTase and RNAse inhibitor were added to the cooled template mixture and incubated for 1 h at 50 °C before an enzyme inactivation step was performed at 70 °C for 15 min. qRT-PCR was performed using SYBR Green PCR Master Mix (Takara Bio Inc., Shiga, Japan) and a 7500 Real-time PCR System (Applied Biosystems, Waltham, MA). Glyceraldehyde-3-phosphate dehydrogenase (GAPDH) was used as an internal control. Complementary DNA samples were diluted 1.5-fold, and qRT-PCT was performed using an ABI-7500 Real-time PCR System (Applied Biosystems, Waltham, MA) with SYBR Premix Ex-Taq II (Takara Bio Inc., Shiga, Japan) according to the manufacturer's instructions. The reactions were performed in a total volume of 20 µl that contained 0.4 mM of each primer (Supplementary Table S2). Each PCR series included a no-template control that contained water instead of cDNA and a reverse transcriptase-negative control for each gene. Triplicate measurements were performed for all reactions. Different samples were evaluated using 96-well plates in the gene expression experiments, and all samples were analyzed on a single plate for the endogenous control experiments. The results were analyzed using the critical threshold (∆C T ) methods in the ABI-7500 software program with the Norm finder and geNorm-plus algorithms. The primers were designed using Primer Express software (Applied Biosystems, Waltham, MA).
Functional annotation and pathways. To functionally annotate the most significant genes, GO analysis was performed using DAVID, version 6.8 20 . GO was analyzed using a modified Fisher's exact P value in the DAVID program. P-values less than 0.001 were considered to be greatly enriched in the annotation category. To determine the possible biological pathways involved in the GSK-J4 treated PM and BV-2 microglial cells, a gene classification analysis of the down-regulated genes was performed using the PANTHER classification system version 9.0 (http://www.pantherdb.org), as described previously 36 . Genes from the datasets that were associated with biological pathways in the PANTHER Pathways Knowledge Base were considered for literature analysis.
Canonical pathway analysis of datasets. An Ingenuity Pathway Analysis (IPA) (Ingenuity Systems, http://www.ingenuity.com, CA) was performed to analyze the most significant canonical pathways in the datasets as previously described 26 . The genes from datasets associated with canonical pathways in the Ingenuity Pathways Knowledge Base (IPAKB) were considered for literary analysis. The significance of the associations between datasets and canonical pathways was measured in the following manner: (1) the ratio of the number of genes from the dataset that mapped to a canonical pathway was divided by the total number of genes that mapped to the same canonical pathway; and (2) Fisher's exact test for a P value indicating the probability that the association could be explained by chance. After uploading the datasets, gene identifiers were mapped to corresponding gene objects, and the genes were overlaid onto a global molecular network in the IPAKB. Gene networks were algorithmically generated based on connectivity.
Transcription factor binding motif enrichment analysis. NCBI reference sequence mRNA accession numbers were subjected to transcription factor binding motif analysis using the web-based software Pscan 25 . The JASPAR 66 database of transcription factor binding sequences was analyzed using enriched groups of −950 base pair (bp) sequences to +50 bp of the 5′ upstream promoters. The range −950 to +50 was selected from the range options in Pscan to obtain the best coverage for a −1000 to +50 bp range.
Statistical analysis. The data were analyzed using Origin Pro 8 software (Origin Lab Corporation, Northampton, MA, USA). Each value is expressed as the mean ± standard error of the mean (SEM). The statistical analyses were performed using SPSS 17.0 software (SPSS Inc., IL, USA). The data were tested using one-way ANOVA, with the exception of the in vivo tests that were analyzed by two-way ANOVA-repeated measures followed by Tukey's HSD post-hoc correction for comparisons. * P < 0.01 and ** P < 0.001 were considered significant.